LBNL Report Number
Solar power is increasingly becoming an important contributor to global electricity systems. Because solar PV power output is both variable and uncertain, there are concerns about how its inclusion in system operations in significant amounts affects conventional power systems operations. Instances of clouds moving over a PV array lead to short-term variability in power output and imperfect forecasts. How utility operations can be changed to more economically integrate large amounts of solar PV power is an open question currently being considered by many utilities.
This report proposes a systematic framework for analyzing implications for operating procedures and corresponding increases in operating costs due to uncertainty and variability in renewable resources. We use the framework to quantify the integration costs and reliability implications associated with sub-hourly solar power variability and uncertainty through a case study based on the projected generation portfolio of a utility in the southwest (Arizona Public Service). We then show how changes in system operations may affect these costs. The costs of sub-hourly solar variability, short-term forecast errors, and day-ahead (DA) forecast errors are estimated as the difference in production costs between a case in which sub-hourly solar variability and uncertainty are fully integrated into the analysis, and the case in which these factors are not considered.
In a lower PV scenario (8.8% of annual demand) we find that the operational challenges are relatively modest and estimate the integration cost to be $1.88/MWh-PV. The integration costs are primarily due to an increase in the cost of balancing reserves held during hour-ahead scheduling. In a high-PV scenario (17% of annual demand), however, we find the operational challenges to be more substantial during low-load and high solar periods. With high PV, we find that curtailments of renewable energy reach a very high level (17.8% of the renewable potential) and that satisfying balancing reserve requirements is challenging in a few hours of the year.
These results are based on several conservative assumptions about system flexibility, including no trade between the utility and its neighbors, no demand response, all operating reserves (both up and down) provided by thermal power plants, several must-run coal plants, and constant output from all nuclear plants. For a high-PV case to be practical, some solution to these operational challenges will be necessary. We include a “flexible nuclear” case as one option for introducing flexibility during low-load and high solar periods. We find that the integration cost drops substantially with increased flexibility. With increased flexibility the estimated integration costs vary between $1.0 and $4.4/MWh-PV in the high PV scenario. Increased flexibility also reduces the curtailment of renewables to between 0.9% and 9.1% of the renewable potential, indicating that the increased system flexibility makes it much easier to absorb high solar PV penetration levels. Other sources of system flexibility, not investigated in this study, are also likely to reduce the integration cost and mitigate operational challenges.